1. Field of the Invention
This invention relates to the optical detection of different chemical species and, particularly, to the identification and/or measurement of chemical substances of interest through detection of changes in the optical properties of a reagent/solvent caused by the reaction with or dissolution of the said chemical substance in the reagent/solvent. More specifically, the present invention is directed to optically detecting the presence, and/or concentration of species which are in the gas or vapor phase or are dissolved in a liquid matrix and, especially, to sensing changes in the optical properties of a confined light conducting medium resulting from exposure of the confining vessel to the sample matrix of interest. Accordingly, the general objects of the present invention are to provide novel and improved techniques of such character.
2. Description of the Prior Art
The use of fiber optics in chemical analysis is known in the art. In the prior art, there are two main classes of fiber optic based sensors employed in such analysis. In a first type of prior art sensor, a relatively long length of optical fiber is deployed as the sensing element and analysis light through the fiber interacts with the medium surrounding the fiber. This interaction occurs because the light used to internally illuminate the fiber, i.e., the analysis light, penetrates the external medium to a distance equivalent to 1/4 of its wavelength, i.e., to a distance much less than the radius of the fiber. Such sensors are called evanescent wave sensors. The main disadvantage incident to the use of evanescent sensors is limited sensitivity. (See for example, "Fiber Optic Optrodes for Chemical Sensing", Brenci and Baldini, in Proceedings, 8th International Conference on Optical Fiber Sensors, pages 313-319, 1992; and "Fiber-Optic pH Sensor Based on Evanescent Wave Absorption Spectroscopy", Ge et al, Analytical Chemistry, volume 65, pages 2335-2338, 1993.)
In the second type of prior fiber optic based chemical sensor, which has found more practical applications when compared to the above-mentioned type device, light launched into the proximal end of an optical fiber emerges at the distal end thereof to interact with a "target" substrate or solution that is affected by the chemical composition of the medium in which the target is present. The light interaction can be probed by observing changes in either light absorption or the emission properties of the target substrate/solution. In a fluorometric mode of operation, a light beam of a wavelength suitable to excite fluorescence is launched through the fiber and the emitted fluorescent light is collected by, for example, the same fiber and separated by a suitable optical arrangement such as a dichroic mirror. (See, e.g., "Enzyme-based Fiber Optic Zinc Biosensor", Thompson and Jones, Analytical Chemistry, volume 65, pages 730-734, 1993); and "Fiber Optic pH Sensor Based on Phase Fluorescence Lifetime", Thompson and Lakowicz, Analytical Chemistry, volume 65, pages 853-856, 1993). Alternatively, the emitted fluorescent light can be collected by a second fiber(s). (See, e.g., "Novel Techniques and Materials for Fiber Optic Chemical Sensing", Wolfbeis, in Optical Fiber Sensors, Springer Proceedings in Physics, Volume 44, pages 416-424, 1989).
Absorptiometric measurements employing this second type of sensor typically implement a bifurcated collection technique, i.e., a second fiber(s) is used to receive the light to be analyzed. Most commonly, in absorption-type sensors, a reflecting optical target containing an immobilized reagent which undergoes a spectral change upon interaction with the analytes of interest in the surrounding medium is located at the fiber tip. Light (monochromatic or broadband) launched through the fiber is reflected off this target and single or multiwavelength measurements are made on the reflected light. (See, e.g., "Potentiometric and Fiber Optic Sensors for pH Based on an Electropolymerized Cobalt Porphyrin"), Blair et al, Analytical Chemistry, volume 65, pages 2155-2158, 1993; "Fiber Optic Sensors for pH and Carbon Dioxide Using a Self Referencing Dye", Parker et al, Analytical Chemistry, volume 65, pages 2329-2334, 1993; and "Current Developments in Optical Biochemical Sensors", Narayanaswamy, Biosensors and Bioelectronics, Volume 6, pages 467-475, 1991). Such immobilized reagents can also be used for fluorescence measurements. (See, e.g.,"Fluorocarbon-based Immobilization of a Fluoroionophore for Preparation of Fiber Optic Sensors", Blair et al, Analytical Chemistry, volume 65, pages 945-947.)
The employment of an immobilized reagent, while attractive in theory, generally results in a sensor with a limited lifetime due to reagent loss from photodecomposition or leaching. To solve this problem, resort has been had to renewing the reagent and, particularly, to flowing the reagent the sensor probe. (See, e.g., "Measurement of Seawater pCO.sub.2 Using a Renewable-Reagent Fiber Optic Sensor with Colorimetric Detection", DeGrandpre, Analytical Chemistry, volume 65, pages 331-337, 1993). In the extant art embodying such flow-through sensors, the tip of the optical fiber is typically located at a first end of a cylindrical chamber, and a reflector is disposed at the opposite end of the chamber. Provision is made for the continuous introduction of a suitable reagent and its withdrawal via conduits which are in fluid communication with the chamber and, typically, oriented in parallel with the optical fiber. The chamber will in part be permeable to the analyte of interest. The analyte thus permeates through a chamber wall and reacts with the reagent thereby producing a change that can be optically monitored. The rate of reagent flow governs the attainable sensitivity, i.e., sensitivity increases with decreasing flow, and response time, i.e., response time decreases with increasing flow. In the reflectance mode, as described above, the effective path length is twice the distance between the fiber tip and the oppositely disposed reflector. The sensor can also be configured with a transmitting fiber at one end of the chamber and an oppositely disposed receiving fiber whereby the path length essentially becomes the length of the chamber. In either case, limited pathlengths are generally attainable, due to severe light loss, and maximum achievable sensitivity is very limited.
Collection of analytes into a reagent flowing through a permeable membrane is well known in the art. The collected analyte is typically measured colorimetrically or fluorometrically in a system external and separate from the collector with or without further reagent addition and reaction. If a sufficient membrane area is provided for analyte collection, parts per trillion levels of analytes can be detected in favorable cases. (See, e.g., "Determination of Gaseous Hydrogen Peroxide at Parts per Trillion Levels with a Nafion membrane Diffusion Scrubber and a Single-Line Flow-Injection System", Dasgupta et al, Analytica Chimica Acta, Volume 260, pages 57-64, 1992; and "Measurement of Atmospheric Ammonia", Dasgupta et al, Environmental Science and Technology, Volume 23, pages 1467-1474, 1989). The sensitivity of a renewable-reagent fiber optic sensor employing this mode of analyte collection would be expected by those skilled in the art to be very low because the length of the membrane that can be used is constrained by the light loss through the membrane.
A renewable reagent liquid core waveguide chemical sensor which uses a membrane material as both the sampling and waveguiding component has been reported. (See, e.g., "Liquid Core Waveguides for Chemical Sensing", Hong and Burgess, Proceedings SPIE, Vol. 2293, pgs. 71-79, 1994). This paper proposes a liquid core waveguide comprised of then available permeable polymers, i.e., PTFE and FEP. Such waveguides, however, have very limited utility because their refractive indexes are greater than that of water, i.e., exceed 1.33, and because they are not optically clear and thus can conduct light for only a short distance. Further, while permeable to gases, such polymers do not have sufficient permeability to function as a sensor probe with significant commercial applicability. In the use of such waveguides, it is necessary to incorporate ethylene glycol or some other non-aqueous solvent(s) in the liquid core, to raise the refractive index of the core above that of the containment tube, and this further limits utility.
The paper discussed immediately above also discussed the use of Teflon A.F. as an internal coating on a porous polyethylene hollow support fiber, the thickness of the coating being on the order of 1 .mu.m. The resulting tube was filled with an acid-base indicator and reportedly responded to relatively high concentrations of ammonia gas (statically deployed in the vapor space over a 0.01 M solution of ammonia in a 50 mL capacity closed vessel) with a seemingly rapid response time (ca. 1 min). Coating the inside of a thin-walled porous fiber with the "AF" solution to produce an ultrathin layer of the latter polymer to produce a reliable or reproducible waveguide is not a commercially practical solution because, as is well known, "Teflon", including the AF variety, has very poor surface adhesion to most material unless special adhesion promoters are used (see, e.g., P. Dress and H. Franke, "A Cylindrical Liquid-Core Waveguide", Applied Physics, Part B, Volume 63, pages 12-19, 1996); and because such surface promoters can compromise the structural integrity of the porous tubing used as support. Also, it is impossible to produce uniform thicknesses of a polymer coating in the manner reported over any reasonable length of a tube and, especially, to produce such a coating in a reproducible manner from one batch to another. Further, it would be impractical to produce such thin coatings on a practical basis without periodic occurrence of pinholes and this would make it impossible to use such tubing in in-vivo physiological applications or in any situation involving significant external pressure, e.g., for a situation in which a sensor is to be immersed in the depths of the ocean. In the latter case, the high collapsibility of supporting porous membrane tubes when pressure is applied from the outside will also compromise the structural integrity of such a sensor.
Importantly, the data provided by Hong and Burgess for the response speed of their "Teflon AF" coated tube teaches away from the use of a polymeric tube solely composed of "Teflon AF", rather than a bilayer structure involving a porous support structure and a "Teflon AF" adlayer. This is because the response time in the Hong-Burgess design is solely due to the permeation through the 1 .mu.m thick "Teflon AF" layer, i.e, the transport in the support structure occurs through the free pore space in a microsecond time scale (the characteristic diffusion time of ammonia a gas with a diffusion coefficient of 0.25 cm.sup.2 /s through a 55 .mu.m deep (see Hong and Burgess, page 78) air-filled pore is only 120 msec. The .about.1 min response time for this device for a 1 .mu.m thick layer as seen in FIG. 9 of Hong and Burgess is actually not fast but very slow, when the thickness is taken into account. It is well known that the characteristic time (loosely, response time for transport) for diffusive or permeative transport across a polymer wall varies directly with the square of the thickness of the polymer wall and inversely as the diffusion coefficient of the analyte of interest through the polymer. For a given analyte and polymer the diffusion coefficient remains constant and thus the response time increases with the square of the thickness (see, e.g., Dasgupta, P. K., "A Diffusion Scrubber for the Collection of Atmospheric Gases", Atmospheric Environment, Volume 18, pages 1593-1599, 1984). Hong and Burgess's own data on PTFE membranes (page 78) show the same behavior. The response time for ammonia decreases by a factor of .about.4 as the membrane thickness is reduced by a factor of two (from 150 to 84 .mu.m). Accordingly, if a polymer of .about.1 .mu.m thickness produces a response time .about.1 min (and this for ammonia, a relatively low molecular weight, small, fast diffusing molecule), the response time for a 75 .mu.m thick tube would be expected to be 5,625 minutes or close to 4 days, if one were to follow the teachings of Hong and Burgess. A "Teflon AF" tube having a wall thickness significantly less than 75 .mu.m would not have sufficient rigidity to be employed as a sensor probe.